Chapter 16 / Analysis of Statically Indeterminate Structures

chap-16 17/1/2005 16: 29 page 467 Chapter 16 / Analysis of Statically Indeterminate Structures Statically indeterminate structures occur more frequently in practice than those that are statically determinate and are generally more economical in that they are stiffer and stronger. For example, a fixed beam carrying a concentrated load at mid-span has a central displacement that is one-quarter of that of a simply supported beam of the same span and carrying the same load, while the maximum bending moment is reduced by half. It follows that a smaller beam section would be required in the fixed beam case, resulting in savings in material. There are, however, disadvantages in the use of this type of beam for, as we saw in Section 13.6, the settling of a support in a fixed beam causes bending moments that are additional to those produced by the loads, a serious problem in areas prone to subsidence. Another disadvantage of statically indeterminate structures is that their analysis requires the calculation of displacements so that their cross-sectional dimensions are required at the outset. The design of such structures therefore becomes a matter of trial and error, whereas the forces in the members of a statically determinate structure are independent of member size. On the other hand, failure of, say, a member in a statically indeterminate frame would not necessarily be catastrophic since alternative load paths would be available, at least temporarily. However, the failure of a member in, say, a statically determinate truss would lead, almost certainly, to a rapid collapse. The choice between statically determinate and statically indeterminate structures depends to a large extent upon the purpose for which a particular structure is required. As we have seen, fixed or continuous beams are adversely affected by support settle- ment so that the insertion of hinges at, say, points of contraflexure would reduce the structure to a statically determinate state and eliminate the problem. This procedure would not be practical in the construction of skeletal structures for high-rise buildings so that these structures are statically indeterminate. Clearly, both types of structures exist in practice so that methods of analysis are required for both statically indeterminate and statically determinate structures. In this chapter we shall examine methods of analysis of different forms of statically indeterminate structures; as a preliminary we shall discuss the basis of the different methods, and investigate methods of determining the degree of statical and kinematic indeterminacy, an essential part of the analytical procedure. 467 chap-16 17/1/2005 16: 29 page 468 468 • Chapter 16 / Analysis of Statically Indeterminate Structures 16.1 FLEXIBILITY AND STIFFNESS METHODS In Section 4.4 we briefly discussed the statical indeterminacy of trusses and established a condition, not always applicable, for a truss to be stable and statically determinate. This condition, which related the number of members and the number of joints, did not involve the support reactions which themselves could be either statically determinate or indeterminate. The condition was therefore one of internal statical determinacy; clearly the determinacy, or otherwise, of the support reactions is one of external statical determinacy. Consider the portal frame shown in Fig. 16.1. The frame carries loads, P and W ,in its own plane so that the system is two-dimensional. Since the vertical members AB and FD of the frame are fixed at A and F, the applied loads will generate a total of six reactions of force and moment as shown. For a two-dimensional system there are three possible equations of statical equilibrium (Eq. (2.10)) so that the frame is externally statically indeterminate to the third degree. The situation is not improved by taking a section through one of the members since this procedure, although eliminating one of the sets of reactive forces, would introduce three internal stress resultants. If, however, three of the support reactions were known or, alternatively, if the three internal stress resultants were known, the remaining three unknowns could be determined from the equations of statical equilibrium and the solution completed. A different situation arises in the simple truss shown in Fig. 4.7(b) where, as we saw, the additional diagonal results in the truss becoming internally statically indeterminate to the first degree; note that the support reactions are statically determinate. In the analysis of statically indeterminate structures two basic methods are employed. In one the structure is reduced to a statically determinate state by employing releases, i.e. by eliminating a sufficient number of unknowns to enable the support reactions and/or the internal stress resultants to be found from a consideration of statical equilibrium. For example, in the frame in Fig. 16.1 the number of support reactions would be reduced to three if one of the supports was pinned and the other was a pinned W B C D P A F RA,H RF, H MA MF FIGURE 16.1 Statical indeterminacy of a portal RA,V RF, V frame chap-16 17/1/2005 16: 29 page 469 16.2 Degree of Statical Indeterminacy • 469 roller support. The same result would be achieved if one support remained fixed and the other support was removed entirely. Also, in the truss in Fig. 4.7(b), removing a diagonal, vertical or horizontal member would result in the truss becoming statically determinate. Releasing a structure in this way would produce displacements that would not otherwise be present. These displacements may be calculated by analysing the released statically determinate structure; the force system required to eliminate them is then obtained, i.e. we are employing a compatibility of displacement condition. This method is generally termed the flexibility or force method; in effect this method was used in the solution of the propped cantilever in Fig. 13.22. The alternative procedure, known as the stiffness or displacement method is analogous to the flexibility method, the major difference being that the unknowns are the displacements at specific points in the structure. Generally the procedure requires a structure to be divided into a number of elements for each of which load–displacement rela- tionships are known. Equations of equilibrium are then written down in terms of the displacements at the element junctions and are solved for the required displacements; the complete solution follows. Both the flexibility and stiffness methods generally result, for practical structures having a high degree of statical indeterminacy, in a large number of simultaneous equations which are most readily solved by computer-based techniques. However, the flexibility method requires the structure to be reduced to a statically determinate state by inserting releases, a procedure requiring some judgement on the part of the analyst. The stiffness method, on the other hand, requires no such judgement to be made and is therefore particularly suitable for automatic computation. Although the practical application of the flexibility and stiffness methods is generally computer based, they are fundamental to ‘hand’ methods of analysis as we shall see. Before investigating these hand methods we shall examine in greater detail the indeterminacy of structures since we shall require the degree of indeterminacy of a structure before, in the case of the flexibility method, the appropriate number of releases can be determined. At the same time the kinematic indeterminacy of a structure is needed to determine the number of constraints that must be applied to render the structure kinematically determinate in the stiffness method. 16.2 DEGREE OF STATICAL INDETERMINACY In some cases the degree of statical indeterminacy of a structure is obvious from inspection. For example, the portal frame in Fig. 16.1 has a degree of external statical indeterminacy of 3, while the truss of Fig. 4.7(b) has a degree of internal statical indeterminacy of 1. However, in many cases, the degree is not obvious and in other cases the internal and external indeterminacies may not be independent so that we need to consider the complete structure, including the support system. A more formal and methodical approach is therefore required. chap-16 17/1/2005 16: 29 page 470 470 • Chapter 16 / Analysis of Statically Indeterminate Structures X X (a) (b) u y x uz u u z y z x z x u FIGURE 16.2 X y u y Statical x indeterminacy of a ring (c) RINGS The simplest approach is to insert constraints in a structure until it becomes a series of completely stiff rings. The statical indeterminacy of a ring is known and hence that of the completely stiff structure. Then by inserting the number of releases required to return the completely stiff structure to its original state, the degree of indeterminacy of the actual structure is found. Consider the single ring shown in Fig. 16.2(a); the ring is in equilibrium in space under the action of a number of forces that are not coplanar. If, say, the ring is cut at some point, X, the cut ends of the ring will be displaced relative to each other as shown in Fig. 16.2(b) since, in effect, the internal forces equilibrating the external forces have been removed. The cut ends of the ring will move relative to each other in up to six possible ways until a new equilibrium position is found, i.e. translationally along the x, y and z axes and rotationally about the x, y and z axes, as shown in Fig. 16.2(c). The ring is now statically determinate and the internal force system at any section may be obtained from simple equilibrium considerations. Torejoin the ends of the ring we require forces and moments proportional to the displacements, i.e.

Load more